1. Field of the Invention
The present invention relates to an mounting system and a method for mounting electronic parts and, more particularly, to a mounting system and a method for mounting QFP (quad-flat-package) semiconductor integrated circuit devices on a printed circuit board, a test socket, or the like.
2. Background of the Invention
Although a electronic part such as semiconductor integrated circuit device is mounted on a mounting body such as a printed circuit board and a test socket by an automatic mounting system, alignment therebetween has become an important problem. That is, a number of leads coming out of the electronic part must be precisely in contact with the corresponding terminals on the board without making short-circuiting any other terminals. In recent years, electronic parts carry thereon leads increasing in number and correspondingly narrower in pitch between them.
Generally, since an electronic part is attracted by an attraction arm and carried and then mounted on a board, there is misalignment between the mutual positions thereof, and hence the correction is needed to precisely align the mutual positions thereamong.
In such circumstances, for example, as disclosed in Japanese Patent Publication No. Hei.3-3959, the differences in position between some selected leads of the electronic part and the corresponding bonding terminals on the p.c. board are detected, and based on the thus-obtained data the misalignment is corrected.
Referring to FIG. 9, the conventional implementation technique will be set forth. The whole optical images of a printed circuit board 320 and an electronic part 322 carried by an attraction arm 321 are formed by cameras 311a, 311b, respectively, and put into a printed circuit board position recognizer 312a and an electronic-part (QFP) position recognizer 312b, respectively, each being stored in the memory address as a set of points in absolute coordinate system.
FIG. 10 shows the plan views of the printed circuit board and the electronic part represented in the X-Y coordinate system. The QFP recognizer 312b selects the three leads 351 to 353 at one ends of the three selected sides, respectively, of the electronic part 322 and stores therein their free-end-center coordinates as the first to third position data 314a to 314c.
On the other hand, the printed circuit board recognizer 312a detects, from the optical image of the printed circuit board 320 sent thereto, the center coordinates of the first to third bonding positions at which the free ends of the first to third leads 351 to 353 are to be bonded, and stores them therein as the data of the first to third bonding positions 313a to 313c.
Next, the data of the first and second lead positions 314a, 314b and the first and second bonding positions 313a, 313b are put into an angle arithmetic section 315 where as the rotation angle of the attraction arm, .theta..sub.A is calculated by the following equation (1) as expressed: EQU .theta..sub.A =tan.sup.-1 {(y.sub.361 -y.sub.362)/(x.sub.361 -x.sub.362)}-tan.sup.-1 {(y.sub.351 -y.sub.352)/(x.sub.351 -x.sub.352)} (1)
where (x.sub.351, y.sub.351) and (x.sub.352, y.sub.352) are the coordinates of the first and second leads, respectively; and (x.sub.361, y.sub.361) and (x.sub.362, y.sub.362) are the coordinates of the first and second bonding positions, respectively.
The obtained angle .theta..sub.A is outputted as it is to the outside and further put into a base-point arithmetic section 316 which computes the coordinate (x.sub.371, y.sub.371) of the first lead when the electronic part has been .theta..sub.A rotated by attraction arm 321, which is used as the first base point. The first base point is calculated, letting (x.sub.c, y.sub.c) be the coordinate of the bonding position and (Dx.sub.1c, Dy.sub.1c) be the differences in X- and Y-directions between the bonding position to the first lead, by the following equation (2): EQU x.sub.371 =x.sub.c -Dx.sub.1c cos .theta..sub.A +Dy.sub.1c sin .theta..sub.A EQU x.sub.371 =y.sub.c -Dy.sub.1c cos .theta..sub.A -Dx.sub.1c sin .theta..sub.A (2)
The coordinates (x.sub.372, y.sub.372) (x.sub.373, y.sub.373) of the second and third base points, respectively, corresponding to the second and third leads are obtained in the same way as the coordinate of the first base point.
The coordinate data 317a, 317b of the first and second base points, and the coordinates 313a, 313b of the first and second bonding positions are put into a Y.sub.A arithmetic section 319. Additionally the coordinate data 317c and 313c of the third base point and the third bonding position are put into an X.sub.A arithmetic section 318. Thus corrective amounts Y.sub.A and X.sub.A are obtained by the following equations (3) and (4): EQU Y.sub.A ={(y.sub.361 -y.sub.371)+(y.sub.362 -y.sub.372)}/2 (3) EQU X.sub.A =x.sub.363 -x.sub.373 (4)
The resulting corrective amounts X.sub.A and Y.sub.A are outputted from the arithmetic sections to the outside world.
Electronic part 322 is adjusted in angle based on the aforesaid angle .theta..sub.A, and in position by corrective amounts X.sub.A and Y.sub.A, and whereby the lead free-ends of the electronic part can be aligned correctly with the corresponding bonding position of the P.C. board.
In the above-mentioned conventional implementation technique for electronic parts, the positional correction is made only the three specified leads of an electronic part to be bonded as base leads, and hence deformation of them within the allowable range may cause a problem that correct positioning for mounting the electronic part is impossible.
FIGS. 11(a) and 11(b) shows as an example an electronic part with all leads having a length of 3 mm and some of them deformed. An implementation-allowable range from the design value between the free-end of the lead and the corresponding bonding position on the P.C. board is set to .+-.0.15 mm. As shown in FIG. 11(a), the first to third selected leads 351b, 353b, and the fourth lead (end lead) 300 at the opposite end to the third lead, get slightly inclined. Furthermore it is assumed that in P.C. board 320 on which the electronic parts are mounted the bonding positions are ideally formed.
FIG. 11(b) represents the inclined selected-leads and their coordinates. The first lead 351b has a coordinate of (110, 43.5) at the root center and an inclination of +0.1 mm in Y-direction. Thus the coordinate of the free-end center is (113, 43.6). The second lead 352b has a coordinate of (110, 43.5) at the root center, and an inclination of -0.1 mm in Y-direction. Thus the coordinate of the free-end center is (87, 43.6). The third lead 353b has a coordinate of (90.5, 57) at the root center and is inclined by -0.1 mm with respect to X-direction. It therefore has a coordinate (90.4, 60) at the free-end center. The fourth lead 300 has a coordinate of (100.5, 57) at the root center, and is inclined by +0.1 mm with respect to X-direction. It therefore has a coordinate (109.6, 60) at the free-end center.
On the other hand, in the printed circuit board 320 the center coordinates of the first to third bonding positions 351b, 352b, 353b, and the fourth bonding position corresponding to the fourth lead 300 are (113, 43.5), (87, 43.5), (90.5, 60), and (109.5, 60), respectively. By substituting the above-mentioned values into the above-mentioned equations (1) to (4), we obtain an corrective amount of angle .theta..sub.A =0.44. Additionally are obtained the first, second, and third base points (112.95, 43.5), (86.95, 43.5), and (90.48, 60.07), respectively; and the position-corrective amount Y.sub.A =0 and X.sub.A =0.02.
All the first to third base points are within the allowable range for implementation of .+-.0.15 mm in both X- and Y-direction with respect to the first to third bonding positions. The free-end center of the fourth lead 300 has a coordinate (109.7, 59.27) after the corrective rotation, and accordingly are largely out of the allowable range for implementation by as much as 0.2 mm in X-direction and 0.73 mm Y-direction with respect to the corresponding bonding position.
The electronic part 322 shown in FIG. 11(b), all the leads of which get deformed within the allowable range, and hence from this viewpoint should be mounted correctly by virtue of implementation system. As described above, however, the conventional implementation technique can not always mount correctly on a P.C. board an electronic part if the leads selected for calculation of the base points and angle get inclined, and this has become the cause of resulting in rejected implementation by the conventional electronic-part implementation system.